GoKawiilApple has just reportedly signed drivers for Nvidia eGPUs, allowing them to finally work on Apple silicon devices. Tiny Corp said in its X post that the company has finally approved the software, letting users pair the GPUs with Macs for AI LLM processing. In fact, the company claims that installing the drivers is now so easy that “a Qwen could do it.” The company first tested an eGPU on Apple Silicon in May 2025, but now that it’s supported by Apple, it means that users no longer have to use workarounds for supported hardware, like disabling System Integrity Protection, to get the system working.
If you have a Thunderbolt or USB4 eGPU and a Mac, today is the day you've been waiting for! Apple finally approved our driver for both AMD and NVIDIA. It's so easy to install now a Qwen could do it, then it can run that Qwen... pic.twitter.com/daUsyBHh1WApril 1, 2026
Tiny Corp is the company behind the tinybox, an AI accelerator built around four high-end GPUs. It was famously known for butting heads with AMD due to driver issues, with AMD CEO Dr. Lisa Su stepping in just to get things right. At the moment, the company is selling the red v2, which is powered by four AMD 9070XTs and costs $12,000, and the green v2 Blackwell, which costs $65,000 and has four RTX Pro 6000 Blackwell GPUs. It’s also planning to launch the exabox in 2027, which will come with 720 RDNA5 AT0 XL GPUs to deliver around 1 exaflop of computing power for around $10 million.
High-end Apple computers recently became popular with the rise of AI agents like OpenClaw. In fact, this frenzy has driven a shortage of Macs that came with massive amounts of Unified Memory, with the delivery window pushed from six days to six weeks. It has even gotten to the point that Cupertino no longer offers the 512GB Unified Memory option for the Mac Studio, while the 256GB model received a $400 price bump.
Article continues below
This custom driver seemingly did not come from the GPU maker, though, with Tiny Corp seemingly working on it on its own. Subsequently, this means that the driver is designed for running AI LLMs and not for gaming, surely disappointing people who don’t want to own two different PCs for work and entertainment. Nevertheless, this is still a game-changer for people working with artificial intelligence, as they could now potentially do training or inference (with some limitations) without needing a dedicated AI supercomputer like Tiny Box.
Follow Tom's Hardware on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.